Physiological Feedback for Real-Time Predictive Model Error Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Predictive models often generate incorrect suggestions, leading to user distrust and reduced usage, and traditional user feedback methods are time-consuming and inefficient.

Innovation Solution

Employing user attention and reaction signals, such as EEG and pupil diameter, to identify when a user perceives an error in a prediction, allowing for real-time correction and model improvement.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional user feedback methods are used to evaluate predictive model accuracy, then model performance can be improved, but the process becomes time-consuming and inefficient

Engineering Contradiction:

Improveprediction accuracy evaluationVSAvoidfeedback collection time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces traditional mechanical feedback methods (manual user input, surveys, explicit ratings) with physiological sensing systems that automatically detect user reactions through biosignals such as EEG, ECG, EMG, and pupillary responses. This substitution enables continuous, passive collection of feedback data without requiring active user participation, thereby resolving the contradiction between measurement precision and time consumption.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service feedback collection by automatically monitoring and analyzing user physiological states without requiring users to manually provide feedback. The computational system processes biosignals in real-time to detect errors and generate feedback signals, eliminating the need for users to spend time explicitly evaluating predictions while maintaining high measurement precision.

Inventive Principle:

Principle #25Self-service

2Speed

If physiological signals are used to detect user reactions, then real-time feedback is enabled, but noisy physiological reactions may lead to false error detections

Engineering Contradiction:

Improvefeedback response timeVSAvoiderror detection accuracy

Core Design Contradiction:

SpeedVSReliability

Solution Approach 1:

The patent merges multiple physiological sensing modalities (EEG, ECG, EMG, pupillary diameter measurement) into a unified feedback system. By combining data from multiple independent biosignal sources, the system achieves more reliable error detection that overcomes the noise and limitations of individual signals, while maintaining real-time responsiveness.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system introduces computational processing and analysis algorithms as intermediaries between raw physiological signals and error detection decisions. These intermediary processing layers filter noise, identify patterns, and correlate multiple biosignals before generating final error detections, thereby improving reliability while preserving real-time feedback capability.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12533064B2Physiological feedback for predictive models

Publication Date: 2026.01.27 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

This document relates to employing biosignals to evaluate predictions made by predictive models. For example, user attention can be inferred from a user attention signal such as gaze. When the user directs attention to a prediction output by a given predictive model, a user reaction signal such as an electroencephalogram or pupillary diameter measurement can be processed to determine whether the user perceives an error. If the user perceives an error, an error indication can be output. Error indications can be used to evaluate the predictive model, replace predictions generated by the predictive model, train the predictive model, etc.